300 SCIENCE AND FRUIT GROWING 



when the tray is perforated, the toxin formed by the plant in it 

 is washed out, and affects the plant in the pot ; when the tray is 

 not perforated, the toxin is retained, and affects the plant in the 

 tray itself. We have here a new light thrown on the reason why 

 good drainage is so essential for the welfare of a plant, namely, 

 to remove the toxin produced by its growth. Up to the present 

 drainage has been regarded as essential owing to its facilitating 

 the access of air to the roots : the roots themselves may want air, 

 and certainly the soil does, but, as has been pointed out already 

 (p. 276), the supply of air which is essential appears to be very 

 small, and any water in the soil would probably contain sufficient 

 for the purpose, so that some further explanation is required 

 of the disastrous consequences following the water-logging of the 

 soil. That it is afforded by the accumulation of plant-toxin was 

 proved by growing mustard in a series of perforated trays, adding 

 to each, at every watering, sufficient water to raise the water 

 contents just short of the maximum which the soil would retain 

 as determined by independent trials and then adding to 

 some of the trays an additional three litres of water, which, 

 of course, drained away. But though this leaching would remove 

 much of the food- mate rial, the plants in these trays did better 

 to the extent of 30 per cent, than those which were not leached, 

 the good done by the removal of the toxin having more than 

 counterbalanced the harm done by the removal of nutrient 

 matter. 



This experiment opens up possibilities for increasing the yield 

 in cases where repeated leaching of the soil is possible, and it 

 further affords one explanation of the great differences which 

 may occur in the action of grass on trees due to differences in the 

 drainage capacity of the soils in question (p. 308) . Work on this 

 subject is still in progress. 



The action of one crop on another being a general one, it 

 follows that trees must affect grass, just as grass affects trees. 

 This was realised in pot experiments wherein grass was grown 

 in the pots, with apple seedlings as a surface crop in the trays. 

 The action was small, because the apple seeds germinated 

 very unsatisfactorily; nevertheless, all the experiments were 

 concordant in showing that an effect did occur (XVII). 



In practice it is difficult to estimate the effect of trees on the 

 grass below them, because there are always other conditions to 

 be taken into consideration, such as the shading effect, and the 

 composite character of the grass, which latter results, when the 



